So I got banned from a subbreddit that uses the name Synth V because my Synth V song had AI art in the video.
It was not even a rule, they made up the rule after my post lol.
People say the same stuff about synth singing and we should be united in this. All tools are on the table when it comes to creativity
So I got banned from a subbreddit that uses the name Synth V because my Synth V song had AI art in the video.
It is this “community” with 1500 members.
I have twice that many subscribers to my youtube channel alone and it has only been going actively now for a few months
It is really not a good business decision to try to block people promoting the product and alienate a very large community, AI artwork is not going anywhere, quite the contrary it is getting larger and larger and I have up until now been uniting these 2 products in my videos ,synth V for the music and Stable Diffusion for the videos are a great combination that the moderators of this subreddit have now taken steps to destroy.
To me its not really an issue of AI being used to generate art, its an issue of training AI on materials not owned by the creators of the AI. With SynthV, the AI voice is created with consent of the singers and all parties make a good faith effort to respect each other’s wishes. AI which produce art such as Dall-e and Stable Diffusion are created largely by scraping the internet for images regardless of the creators consent or not. I believe this is a legal grey area, where using copyrighted works in training an AI is allowed by fair use, which would be used as a legal defense in litigation, not an blanket law. Morally, I think its kind of messed up to use other people work against their consent.
This is all to say that SynthV’s use of AI and other AI are not comparable just by their use of AI.
Also Dreamtonics has no control over a subreddit and doesn’t get involved with fan communities lmao.
It is not a grey area because I could legally right now take 20 living artists and make a collage without AI with their work
and use it without consent if it has changed enough from the source material. This is legally known as fair use. There are a lot of cover songs on that forum where mine are all 100% original. Did every person get consent from the artist before doing a cover song? My finished pieces are also extremely different from any artists work, much more so than a cover song is.What they should have done instead of ban me is embrace me and welcome the cross promotion but like you said they are not affiliated with the company so they do not care about how well it does and flat out stated so in a reply to me. AI is not going anywhere it is already a huge community that is lots bigger than the Synth V community and will only get bigger.
Here is an example of a recent fair use case where someone used the AI art to create a comic. This is no different from me using it as part of a music video. Really though this precedent was set a long time ago, many artists have done collage kind of work that is legal because it was changed enough from the source material.
Again Fair Use is a legal defense, to use your collage example, you could make a collage without AI using art from other artists, but you could be liable to legal action and in a court of law you would need to defend your use of their art by proving it was fair use. Cover songs on that forum are the same, and often are subject to copyright claims from the copyright owners, but also many artists in the Vsynth space encourage/consent to covers providing song stems and untuned files.
The article you cited doesn’t really relate to fair use, as there was not legal dispute that needed to be settled by a fair use defense. However, It does show that AI art can receive a copyright registration and to be fair I don’t seem to see any copyrighted material in that work. With that said copyright registration isn’t the end all be all, plenty of people have managed to copyright things which are absolutely copyright infringement (Sonichu for example). I’ll touch on how this is a concern for AI in my next point.
As far as a “legal grey area”, I say this because in the United States Fair use is the legal defense for using copyrighted material, but not a concrete law that says “You may use copyrighted content for training AI”. Infact as you state AI is a growing technology and I worry that without more self regulation on how the data used to train these models is had, we may see more regulation from governments to protect large copyright holders for example Disney deciding they don’t like their content being used in this way and lobbying to make copyright laws even stricter. I don’t think that’s out of the question when you look at other laws that companies like Disney have had a part in getting passed. People tend to forget that the law is not unchanging, and one sweeping law could very well prohibit all AI created works from copyright registration, and revoke copyright registration to all AI works. Even if we disregard AI, we can clearly see many attempts by lawmakers tying to restrict fair use, with laws attempting to revoke safe harbor for content providers, so I don’t think what I’m saying is such a stretch.
So with all that, I worry that the violation of consent could result in strict regulation for the future of AI. It’s not that I think copyrighted works shouldn’t be used in educational purposes, truly AI like Stable Diffusion are very impressive, and using copyrighted material helps to engineer the foundations of the ML algorithm but releasing ML models which use those copyrighted materials to the public with seemingly no restrictions to me is shooting themselves in the foot. And it’s not like we can’t get incredible results without using materials of copyright owners, SynthV is developed by Dreamtonics and has high quality results without “stepping on anyone’s toes”, and as you note VSynth is a very small piece of a larger AI community, so its like “If the little guys can do it right, why can’t the big guys?”
So if the moral argument doesn’t appeal to you, maybe consider how trillion-dollar copyright holders could respond to the rampant rise of AI technology and decide to create regulation if the community doesn’t regulate itself.
There are already cases that have decided this. It really just boils down to one simple thing, does the created work change the original work enough to be considered fair use but I see your point about the model training. You have to also take into consideration that there are lots and lots of dead artists and lots of material that is in the public domain that was used. Also there are artists, like Beeple, that like AI. He even does contests on twitter sometimes for people to re-create his work with AI. This is one thing, a lot of these outraged people , like the mods in the reddit forum, are probably not even artists themselves but act as if they speak for every artist on the planet. There was quite a scare in the art community when Photography was invented and many of the discussions today parallel that but photography and oil painting are now co-existing and this will also be the case with AI. Really it already is, lots of people are using it in conjunction with other programs. Like I have videos where I use input videos for the base, public domain videos from pexels so the AI is not always the end product it is just one more element, it is just a tool. So in the same way that you cannot copyright a sentence, but can copyright a book with the sentence in it the work that is produced does still have the human hand in it at least for more serious projects. Ty for the reply and the conversation, my main point is just that they are really attacking something that has nothing to do with Synth V and the claim that AI art harms Synth V is outlandish, I have never seen anyone say they are not going to get Synth V because of AI art. lol There are however quite a few people that are anti- synth vocals and this has been the case way before AI art started getting popular.
It looks like you’re using old Reddit. For some reason, the rules are accessible only when you’re using new Reddit, so that’s why you might not have seen them.
No they actually told me in a response that they added the rule because of my video. Lol, you know I do not mind being shut out of the forum,it is very small and I get way more clicks from the AI reddits that have over 100k people compared to the 1500 in their reddit. I really only post in there to get people interested in the AI stuff and I have posted my stuff there before with no issues.It is more about them falsely claiming they represent the Synth V community and speak for Synth V users. I have used this for a very long time and my videos bring attention to Synth V, lots of people like these AI videos and it is really a great match considering many solo artists like myself use Synth V and with the AI we can custom make videos for our songs as well and I have been promoting it because I think within 5 years synth vocals are going to be very common. So to shut out people because they use AI programs makes absolutely zero sense, it is a bigger community and they should be trying to draw people in not kick them out based on their own personal bias towards the program. Here is the reply they sent when I said I did not see the rule, they are just terrible mods.No good mods kick people out for a rule that is not even posted and they are really not very smart to be turning people away that are doing things in a community 100x bigger than theirs. I have more subscribers to my channel then they have reddit and I am not alone in this,there are other people doing AI art in conjunction with Synth V and the AI art is not going away, they are fighting a lost battle.
Yeah, in that case, I’d be OK with adding that rule and removing your post, but banning you by a rule that didn’t exist when you posted it is just unfair.
Great thread. Its good to be around people who understand this world.
Question: Do you believe the kind of behavior you display within these walls of text is very good…?
Anyways, let’s speak about my 2 Cents according to the topic.
AI-generated art has indeed some Pros such as giving artists inspiration (poses, colors, etc.) or even produce backgrounds for personal sketches / drawings when one only feels like creating characters, yes.
But there’s also a HUGE downside to these pictures.
(And I don’t mean the problem of those AIs which work based on stuff swimming around the Internet.)
You know how it might be called when you use an AI-generated piece a la 1:1 …?
Or, in other words, doing so gives the vibes of: “Why learn drawing myself, when the AI does EVERY SINGLE STEP for me?”
If you yourself cannot draw and instead ask another to do the job, then it’s that person who will invest all the effort (for you as a friend ooor for your money).
》But when all it took of your (or another’s) time / skills / resources was typing a few words…?
(In case you use an AI which solely works with your own creations - or stuff an artist allowed you to rely on for making the program smarter - there is of course a valid excuse for being upset.)
If you insist on going a route that saves your time, maybe consider using pictures of 3D-models instead of AI-generated art?
First off, I put a lot more time into my song releases than most people, I do not display a single frame of an anime character, my songs are all original ,not covers, and I record the guitars, the music and mix and master them myself.The art is but a small part of the whole. I have been a digital artist for 20 years and always use whatever tools I feel like and this is what good artists do, Are YOU an artist? Because most of you people complaining about it are not. Small Idea, maybe consider not attacking people that are trying to build the Synth V community because it has definitely not caught on in the states and will not as long as the majority of it is pop covers that only have a still anime image as the logo.
I spend WAY way more time on my stuff than anybody on that reddit.
This is why my videos actually generate views.
It also includes animated lipsynching which also requires some work, you have the audacity to call me lazy?? Show me your work and show me where you do more than me, I will be glad to compare workflows any day any time
I do draw, I do all kinds of art, do you?? Maybe let the artists who are so offended by this speak for themselves, what people like you really have is just a severe case of holier than thou syndrome.
You cannot touch what I do
And to ban someone that for a rule that is not even posted is just amateur hour lmao. This is the thing, if they had that rule posted I would have obliged, these people are trying to push an agenda and when we come check back in 2 years, the AI art will continue to grow and evolve, it will not go away and they are fighting a losing battle and for what purpose? Because they are artists protecting work? If it was stolen,or copyrighted they would be taken down, Youtube is very quick to remove stuff like that, they do not know what they are talking about and the only thing they accomplish by banning people like me is less members and less promotion for synth V.This is my final word on the matter, I am way to busy to debate every non artist that all of a sudden thinks they are the megaphone for artists, and you are not, many artists , like me, use AI and many who are much better and have a wider audience than me, like Beeple.
So do your own research and I will be glad to accept your apology
Frankly if you haven’t seen the immense number of artists speak out against both NFTs and AI art generators, you simply haven’t been paying attention. You’re not going to convince anybody by feigning ignorance.
First of all, you do not speak for anyone but yourself yet you always try to frame your statements as if you represent a multitude of people, such arrogance. You are also proving my point that most of the rabid anti-ai frothing at the mouth full of righteous fury are not artists at all but simply people,like yourself, that seem to revel in cloaking yourself in a very tired holier than thou attitude. You are the one that is ignorant as you have already shown many times and many times I have completely de-bunked your lies, like with your absolute lack of knowledge of the legal status of this tool. There are artists on both sides of the issue and I am far from ignorant, since I have been one of the artists on the pro side I have already at this time engaged in numerous discussions and unlike you I acknowledge both sides and could even argue your side better than you could since you do not. First of all there are very few artists who are affected by this, most of the people complaining are like you, self appointed righteous crusaders that are not even artists, or little known artists that blame their own failures on whatever they can. The fact is unless you are a very well known artist you will not be affected by this at all and you have to be well known for the AI to be able to copy your style (unlike the misinformation out there AI does not copy and paste outlines of art in a collage like fashion) First of all with that, Style has never been something you could copyright and still is not. Many of the artists that it can do this with, are dead, like Van Gogh, H.R. Giger and AI is not doing anything that has not been done before it just does it faster, if artists did not use reference material and draw upon other artists as inspiration all we would have is cave paintings. Now having said that, there are some living artists with legitimate concerns (You are not among them) You have yet to mention Greg Rutkowski, which shows how woefully ignorant you are on the subject since he makes probably the best case against AI. His name was included in some of the very early AI programs ( A lighthouse on a shore by Greg Rutkowski) so his name got used a lot and he also has a style that lends itself to many different kinds of images both fantasy and realistic. Some of his concerns are people searching for his art now might instead find a multitude of AI images and this is definitely a legitimate concern. On the other side of the coin he has generated a lot of publicity due to his popularity but this does not diminish his concerns. If you were actually less ignorant of the subject you would bring up things like this, but all you have ever shown is that you like to rant and rave and will lie and mis represent facts.
Here is the Greg Rutkowski article, I suggest in the future on your losing crusade(and you will lose the matter has already been decided in courts and AI is here to stay, legally and otherwise) you use articles like this and try to actually present facts instead of lies
This artist is dominating AI-generated art. And he’s not happy about it. | MIT Technology Review
Now to show that what I said is correct, here are some pro AI artists
Beeple-707.4k followers and often encourages people to use AI to duplicate his work in contests and winners receive a print.
(1) beeple (@beeple) / Twitter
ergojosh and there are numerous examples of Digital artists
that use AI as a starting point(like me) and than add their own skills to the mix. Here is just one of them but you can easily find many pro AI artists on youtube and elsewhere, you are just blind because you have an agenda. 676k subscribers are on this channel, (which absolutely dwarfs that subreddit with the terrible mods)
Now I am done responding to you period because it is very tiresome to have to de-bunk your messages that are always full of misrepresentation of facts, or flat out lies. I wish you well on your endeavors but your opinion (which again is yours only) on anything does not interest me, I swim in a much bigger lake than you and have no time to fact check all of your messages, I have shown you lie and mis represent and will not do so again because people like you will not change your ways, you enjoy playing the crusader far too much to let go of the role even when the facts show how wrong you are. I will not respond or acknowledge you or even read your reply, I am only interested in people that promote Synth V, not people that are reducing access to it based on their own personal agenda like you and those amateur mods(of which you probably are one you definitely talk like them. )
It has been explained to you multiple times that the primary issue is the non-consensual collection of the training dataset. Your refusal to acknowledge that problem is very telling.